The specific aims of this component are to investigate potential therapeutic medications to prolong the disability-free period of individuals with HIV infection. Previous work at the Scripps HIV Dementia Research Center and the HIV Neurobehavioral Research Center at UCSD identified fatigue as a major cause of HIV-related disability and unemployment. Fatigue and sleep disruption are invariably found in early HIV infection, when individuals remain medically and cognitively healthy, prior to disease progression to AIDS. The focus of this proposal is this early fatigue-related disability and its pathophysiology. Fatigue and sleep disturbance worsen the effects of HIV dementia and hasten both deterioration in quality of life and socio-economic decline. We propose brief trials to test hypotheses relating to fatigue pathophysiology and pharmacological interventions based on reversing pathophysiology. Using a parallel-groups, placebo-controlled, single-blind design, we will rapidly screen medications in HIV-infected subjects for their potential efficacy in controlling disability due to fatigue. Subject involvement with this project will be brief (8 weeks) and only effects related to fatigue are under study. The first hypothesis tested is that the daytime fatigue and increased slow wave sleep found in early HIV infection are mediated by TNF alpha, a fatigue-promoting peptide elevated in HIV. Pentoxifylline inhibits TNF alpha mRNA expression, and may provide a specific intervention against HIV-related fatigue. We will assess the effect of pentoxifylline on fatigue, on levels of TNF alpha throughout the 24 hour day, on waking and sleep EEG recorded through the 24 hour day, on neurocognitive performance, on CANTAB, and on abnormal evoked brain potentials. Sleep will be evaluated both by nocturnal polysomnography and quantitative EEG techniques. Pentoxifylline effects will be compared with placebo and with a non-specific neurostimulant d-amphetamine. THE SECOND HYPOTHESIS tested is the the fatigue and other effects are related to NMDA-receptor-mediated glutamate toxicity, amantadine protected neurons against this toxicity. It afforded less protection to neurons with non-NIMDA-receptor-mediated glutamate toxicity. Amantadine will be tested for its effect on fatigue and the other measures of CNS integrity used in this project. FURTHER HYPOTHESES will be formulated and tested using one or more drugs suggested as promising agents for early clinical trial by other studies within the TSRI HIV Dementia Center's preclinical studies groups working with feline immunodeficiency virus, simian immunodeficiency virus, and transgenic mouse models. We will also STREAMLINE AND SIMPLIFY ASSESSMENT OF HIV-related fatigue and CNS effects by evaluating each of the following 4 techniques for detection of early HIV-related CNS change; (1)psychomotor vigilance test and driving simulator; (2) performance on a code substitution task and a mannequin orientation task; (3) auditory evoked potentials; and (4) P300 event-related potentials.